Interference across time: dissociating short from long temporal interference

Abstract

Our ability to identify an object is often impaired by the presence of preceding and/or succeeding task-irrelevant items. Understanding this temporal interference is critical for any theoretical account of interference across time and for minimizing its detrimental effects. Therefore, we used the same sequences of 3 orientation items, orientation estimation task, and computational models, to examine temporal interference over both short (<150 ms; visual masking) and long (175-475 ms; temporal crowding) intervals. We further examined how inter-item similarity modifies these different instances of temporal interference. Qualitatively different results emerged for interference of different scales. Interference over long intervals mainly degraded the precision of the target encoding while interference over short intervals mainly affected the signal-to-noise ratio. Although both interference instances modulated substitution errors (reporting a wrong item) and were alleviated with dissimilar items, their characteristics were markedly disparate. These findings suggest that different mechanisms mediate temporal interference of different scales.

Keywords: interference; mixture-model analysis; similarity; temporal crowding; visual masking.

Figure 1

An example of a single trial in the masked condition of Experiment 1. There were five possible target-distractor SOAs (40, 60, 80, 100, and 120 ms). The SOA was fixed within a trial but varied between trials. In the unmasked condition, only the target appeared. The task was to reproduce the target’s orientation.

Figure 2

Model comparison in Experiments 1–3. The difference in AICc values between each tested model and the two-misreport mixture model. The two-misreport mixture model had the lowest AICc values (i.e., all differences were positive) in all three experiments.

Figure 3

Mean error distributions and model fits (in black) for the various conditions of Experiments 1–3.

Figure 4

Target report rate (A) and the estimated parameters (B) sd, (C) g, (D) β₁, (E) β₂ as a function of SOA in the masking condition of Experiment 1. For comparison we also plot the unmasked condition (UM). Error bars represent one standard error.

Figure 5

A piecewise regression model fitted to the estimated parameters in the masked/crowded conditions of current Experiment 1 (SOA range: 40–120 ms) and Experiment 2 of Tkacz-Domb and Yeshurun (2021; SOA range:170–470): (A) Target report rate; (B) sd; (C) g; (D) β₁; (E) β₂. The dotted line shows the breakpoint. Error bars represent one standard error. The shaded region corresponds to 95% CIs.

Figure 6

Estimated slopes and 95% CIs (error bars) for the 3 SOA segments generated by the piecewise regression model for the: (A) Target report rate; (B) g parameter (see text).

Figure 7

Target report rate (A) and the estimated parameters (B) sd, (C) g, (D) β₁, (E) β₂ as a function of SOA in the crowded condition, and the uncrowded condition (UC) in: Experiment 2 of the current study where stimulus luminance varied randomly (green), Experiment 1 of Tkacz-Domb and Yeshurun (2021) where all stimuli had similar luminance (blue), and Experiment 3 of Tkacz-Domb and Yeshurun (2021) where stimulus luminance was fixed throughout the experiment (red).

Figure 8

Target report rate (A) and the estimated parameters (B) sd, (C) g, (D) β₁, (E) β₂ as a function of SOA in the masked condition, and the unmasked condition (UM) in Experiment 1 (Mask-Similar; blue), and Experiment 3 (Mask-Dissimilar; green).